Autoregressive Models with Mixture of Scale Mixtures of Gaussian Innovations

Abstract

This paper presents a theoretical and empirical study of likelihood inference for the autoregressive models with finite (m-component) mixture of scale mixtures of normal (Gaussian) (SMN) innovations. This model involves autoregressive models with single and mixture component of innovations, which are frequently used in time series data analysis. An EM-type algorithm for the maximum likelihood estimation is developed and the observed information matrix is obtained. The performance of the proposed model through a simulation study is also evaluated. The model is then applied on a real time series data set.